Abstract:The automatic discovery of a model to represent the history of encounters of a group of patients with the healthcare system -- the so-called ``pathway of patients'' -- is a new field of research that supports clinical and organisational decisions to improve the quality and efficiency of the treatment provided. The pathways of patients with chronic conditions tend to vary significantly from one person to another, have repetitive tasks, and demand the analysis of multiple perspectives (interventions, diagnoses, medical specialities, among others) influencing the results. Therefore, modelling and mining those pathways is still a challenging task. In this work, we propose a framework comprising: (i) a pathway model based on a multi-aspect graph, (ii) a novel dissimilarity measurement to compare pathways taking the elapsed time into account, and (iii) a mining method based on traditional centrality measures to discover the most relevant steps of the pathways. We evaluated the framework using the study cases of pregnancy and diabetes, which revealed its usefulness in finding clusters of similar pathways, representing them in an easy-to-interpret way, and highlighting the most significant patterns according to multiple perspectives.
Abstract:When mining large datasets in order to predict new data, limitations of the principles behind statistical machine learning pose a serious challenge not only to the Big Data deluge, but also to the traditional assumptions that data generating processes are biased toward low algorithmic complexity. Even when one assumes an underlying algorithmic-informational bias toward simplicity in finite dataset generators, we show that fully automated, with or without access to pseudo-random generators, computable learning algorithms, in particular those of statistical nature used in current approaches to machine learning (including deep learning), can always be deceived, naturally or artificially, by sufficiently large datasets. In particular, we demonstrate that, for every finite learning algorithm, there is a sufficiently large dataset size above which the algorithmic probability of an unpredictable deceiver is an upper bound (up to a multiplicative constant that only depends on the learning algorithm) for the algorithmic probability of any other larger dataset. In other words, very large and complex datasets are as likely to deceive learning algorithms into a "simplicity bubble" as any other particular dataset. These deceiving datasets guarantee that any prediction will diverge from the high-algorithmic-complexity globally optimal solution while converging toward the low-algorithmic-complexity locally optimal solution. We discuss the framework and empirical conditions for circumventing this deceptive phenomenon, moving away from statistical machine learning towards a stronger type of machine learning based on, or motivated by, the intrinsic power of algorithmic information theory and computability theory.